Adding the μ-opioid receptor agonist remifentanil to agents used to induce general anaesthesia in electroconvulsive therapy (ECT) can reduce the required doses of induction agents and their unfavourable effects on seizure threshold and quality. However, whether remifentanil has favourable long-term treatment effects in terms of response and remission rates, speed of response and remission, and side-effects has not been studied.This retrospective, register-based cohort study involved patients with major depression consecutively treated at two units at different hospitals in Norway with the same ECT procedure. Both units used thiopental for ECT anaesthesia, but only one unit added remifentanil (R+; n=47; 541 sessions), whereas the other did not (R-; n=119; 1166 sessions). A Cox proportional hazards model for interval-censored data was conducted to examine the effects of remifentanil on the time to response and remission from depressive symptoms, whilst adjusting for age, sex, and baseline depression score.Both R+ and R- patients showed substantial reductions of depressive symptoms, with no difference in the response (76% in both groups) or remission (63% vs 65%) rate. However, R+ patients responded (hazard ratio=0.59; 95% confidence interval: 0.4-0.8) and remitted (hazard ratio=0.72; 95% confidence interval: 0.5-1.0) more slowly, and reported more often side-effects of nausea (30% vs 8%; P0.001), dizziness (22% vs 8%; P=0.027), and headache (48% vs 23%; P=0.004).The use of adjunctive remifentanil was associated with more short-term side-effects and no favourable long-term clinical outcomes. The practice of routinely adding remifentanil to barbiturate anaesthesia should therefore be reconsidered.
